How is Bard able to use current news/data more quickly than ChatGPT?

Question

It seems like ChatGPT has quite a large lag from their training set to when it goes to production, with no updates in between. But with Google Bard it seems like they are able to update their existing models for current events? What makes this possible and why can't ChatGPT do the same?ChatGPT:> what was the last fed rate hike?> As an AI language model, my training cutoff date was in September 2021 and I do not have access to real-time information. However, I can tell you that the Federal Reserve's last interest rate hike occurred on December 16, 2020. At that time, the Federal Open Market Committee (FOMC) raised the target range for the federal funds rate by 0.25 percentage points to a range of 0.25% to 0.50%. This was the first and only interest rate hike of 2020.Bard:> what was the last fed rate hike?> The Federal Reserve raised interest rates by 25 basis points on March 22, 2023, to a target range of 4.75-5%. This was the ninth consecutive increase in interest rates since the central bank began the current rate-raising cycle to combat rising prices.

relatedtitle · Accepted Answer

I assume the model is probably generating a search query, then the results get inserted as context and it uses that to formulate a response.

Mizoguchi · Answer

Could this be a parallelized two step process?
One a Google web search then filling out placeholders in the actual model's response?
I did notice sometimes with Bard the numbers are incorrect when asking for multiple things at a time, say a list of events in a particular city next month.
Bard will add the wrong date to say a hockey game in the returned list but if you ask for the event alone then it will give you the right answer

ldjkfkdsjnv · Answer

Probably its doing a google search then passing it to bard in the prompt